Dr. Arnaud Rosier PhD, recently published an article on how partnering AI with Remote Patient Monitoring (RPM) can reduce innate bias. These bias can be found in the algorithms used in healthcare AI technology. The article was highlighted in Healthcare IT Today.
Studies have shown that AI algorithms that use poorly representative datasets can exhibit signs of bias in their findings. This bias may lead to ethnic, gender, and social discrimination. Unintentional biases combined with personal bias of healthcare professionals may have serious consequences.
One key for reducing the bias of both AI programming and/or healthcare personnel is Remote Patient Monitoring. RPM eliminates bias in data reporting. Because RPM provides data without any influence of human bias. RPM can also add more diversity of datasets used to train AI algorithms.
Participants in most medical studies tend to represent demographics who have more financial and familial resources. Those who have the time to share health readings and/or a support network to help them participate in the medical study, does not fairly represent all patients with those medical conditions. RPM removes helps to remove the biases.
One area where this progress is already occurring is in the area of remote cardiac monitoring.
Dr. Rosier, as CEO of the company Implicity, showed how his company worked with the French government to create a Health Data Hub. Data collected from remote cardiac monitoring devices, created a dataset from various regions and population groups in France. The innovative algorithm created by the AI element of the Health Data Hub is capable of accurately predicting instances of acute heart failure in patients, without regard to their social or ethnic background.
The data was collected from a very wide group of individuals from the economically challenged to the affluent, from inner cities to remote rural communities. Additionally, the information gathered from RPM can be added to aggregate data of historically under-representative demographics.
RPM is also changing how medical research is performed by broadening patient access to studies. Travel time and expenditure prevented many potential participants. By equipping cardiac patients with RPM devices in their homes, patients no longer have to go to a clinic for routine checks of their vitals.
RPM can also help AI become smarter, more adaptable. AI algorithms are used in RPM devices to perform their basic functions. By providing developers with the experiences of actual patients not in tightly controlled research settings, the bias of research study is eliminated. With non-bias RPM readings, AI programs can offer course corrections. AI makes RPM work, and RPM makes AI better.
As AI becomes more utilized, there must be an investment in obtaining and using the data of diverse patients representing different demographics, ages, ethnicities, and regions. RPM clearly has an advantage in obtaining unbiased patient data on a consistent and large-scale basis.
By having your patients participate in a RPM program, like one provided by Medek RPM. You will improve their health outcomes, add to the financial health of your practice, and you are also helping to make medical AI better and free of more biases. Once again, by doing something good for you, you can also be doing something that is good for all of us.
Medek RPM is a leader in the United States for helping medical practices develop and operate remote patient monitoring. Connect with a Medek representative to learn your practice, your patients and your society can benefit.
About Dr. Arnaud Rosier
Dr. Arnaud Rosier, who holds a PhD in symbolic artificial intelligence, is a cardiac electrophysiologist and the founder as well as CEO of Implicity.
